CS 528 Mobile and Ubiquitous ComputingLecture 1a: Introduction
Emmanuel Agu
About Me
A Little about me
WPI Computer Science Professor
Research interests: • mobile computing especially mobile health, computer graphics
Started working in mobile computing, wireless in grad school
CS + ECE background (Hardware + software)
• Current active research: Mobile health apps
• E.g: AlcoGait app to detect how drunk Smartphone owner is
• https://www.youtube.com/watch?v=pwZaoKmfq8c
Administrivia
Administrivia: Schedule
Week 1-8: I will introduce class, concepts, Android (Students: Android programming, assigned projects) Goal: Students acquire basic Android programming skills to do excellent project
Programming apps that use mobile & ubicomp components
Week 9: Students will present final project proposal
Week 9-14: Students work on final project
Week 11: Students present on new mobile APIs, components E.g. machine learning in Android, Augmented Reality
Week 14: Students present + submit final projects
Quizzes (5) throughout
Special Notes: This online offering
Today’s class recorded, video posted to canvas after class
From lecture 2 on: Videos posted days BEFORE class
Class: quick summary of key points, more interactive (Question and Answer), Quiz
Default: I’ll assume all students can make it to class for quizzes
Please email me if you cannot. E.g. different time zone, illness, etc
Requirements to get a Grade
Grading policy:
Presentation (tech topic) 15%, Assigned Projects 35%, Final project: 30%, Quizzes: 20%
Final project phases: (See class website for deadlines)1. Pick partners, form project groups of 5 members
2. Submit 1-slide of proposed idea (problem + envisioned solution)
3. Present project proposal
+ plus submit proposal (intro + related work + methodology/design + proposed project plan)
4. Build app, evaluate, experiment, analyze results
5. Present results + submit final paper (in week 14)
Degree of difficulty of project taken into account in grading rubric
Course Texts Android Texts:
Head First Android Dev, (2nd ed), Dawn and David Griffiths, O'Reilly, 2017
Android Programming: The Big Nerd Ranch (Third edition), Bill Phillips, Chris Stewart and Kristin Marsicano, The Big Nerd Ranch, 2017
Will also use official Google Android documentation
Learn from research papers: Why not text?
Gentle, visual introBootcamp
TutorialVisual kotlin intro
Grader
Will be hired
Class in 2 Halves
2 Halves: About 1 hour 15 mins each half
Break of about 15 mins
Talk to me at the end of class NOT during break
I need a break too
Poll Question
How many students:
1. Own recent Android phones (running Android 4.4, 5, 6 , 7, 8 or 9?)
2. Can borrow Android phones for projects (e.g. from friend/spouse)?
3. Do not own and cannot borrow Android phones for projects?
4. Cannot come to class (e.g. in very different timezone?) Other constraints?
Mobile Devices
Mobile Devices
Smartphones (Blackberry, iPhone, Android, etc)
Tablets (iPad, etc)
Laptops
Smartwatches
SmartPhone Hardware
Smartphones have capabilities beyond calling and texting (or feature phones)
Communication: Talk, SMS, chat, Internet access
Computing: Powerful processors, programmable operating system, Java apps, JVM, apps
Sensors: Camera, video, location, temperature, heart rate sensor, etc
Example: Google Pixel XL 3 phone: 8 core 2.5 GHz/1.6GHz kryo CPU, Adreno 630 GPU, 128GB RAM A PC in your pocket!!
Multi-core CPU, GPU, over 20 sensors (10 hardware sensors, over 10 soft sensors)
Linux OS, JVM, runs OpenGL ES, OpenCL and now Deep learning (Tensorflow)
Communication Computing Sensors+ +Smart =
Qualcomm SnapDragon System on a Chip (SoC)
Core of most high end smartphones shipped in 2020
SoC: Chip that integrates most computer components: CPU, GPU, memory, I/O, storage
Ref: https://arstechnica.com/gadgets/2019/12/qualcomms-new-snapdragon-865-is-25-faster-comes-with-mandatory-5g/
Smartphone Sensors Typical smartphone sensors today
accelerometer, compass, GPS, microphone, camera, proximity
Can sense physical world, inputs to intelligent sensing apps E.g. Automatically turn off smartphone ringer when user walks into a class
Growth of Smartphone Sensors
Smartphone generations have more and more sensors!!
Image Credit: Qualcomm
Future sensors?
• Complex activity sensor,
• Pollution sensor,
• etc
Wireless Networks
Wireless Network Types
Wi-Fi (802.11): (e.g. Starbucks Wi-Fi)
Cellular networks: (e.g. T-Mobile network)
Bluetooth: (e.g. car headset)
Near Field Communications (NFC)
e.g. Mobile pay: swipe phone at dunkin donut
Wi-Fi NFC
Bluetooth
Wireless Networks Comparion
Network Type Speed Range Power Common Use
WLAN 600 Mbps 45 m –90 m
100 mW Internet.
LTE (4G) 5-12 Mbps 35km 120 – 300 mW Mobile Internet
3G 2 Mbps 35km 3 mW Mobile Internet
Bluetooth 1 – 3 Mbps 100 m 1 W Headsets, audio streaming.
Bluetooth LE 1 Mbps 100+ m .01–.5 W Wearables, fitness.
NFC 400 kbps 20 cm 200 mW Mobile Payments
Table credit: Nirjoin, UNC
Different speeds, range, power, uses, etc
Mobile Computing
Mobile Computing
• Human computes while moving• Continuous network connectivity,
• Points of connection (e.g. cell towers, WiFi access point) might change
• Note: Human initiates all activity, (e.g launches apps)
• Wireless Network is passive
• Example: Using foursquare.com on Smartphone
Related Concept: Location-Awareness
Mobile computing = computing while location changes
Location-aware: Location must be one of app/program’s inputs Different user location = different output (e.g. maps)
E.g. User in California gets different map from user in Boston
Program/app
Inputs
Output
Program/app
Inputs
Output
Location
Non-mobile app Mobile app
Location-Aware Example
Location-aware app must have different behavior/output for different locations
Example: Mobile yelp
Example search: Find Indian restaurant
App checks user’s location
Indian restaurants close to user’s location are returned
Example of Truly Mobile App: Word Lens
Translates signs in foreign Language
Location-dependent because location of sign, language? Varies
Acquired by Google in 2015, now part of Google Translate
Some Mobile apps are not Location-Aware
If output does not change as location changes, not location-aware
Apps run on mobile phone just for convenience
Examples:
Distinction can be fuzzy. E.g. Banking app may display nearest locations
Diet recording appMobile banking app
Which of these apps are Location-Aware?
a. Yahoo mail mobileb. Uber app
Notable: Sharing Economy Apps Idea: Share resource, maximize under-utilized capacity
E.g. Uber: share care, Airbnb: Share house
Question: How is mobile/ubicomp used in sharing apps?
Mobile Device Issue: Energy Efficiency Most resources increasing exponentially except battery energy (ref. Starner, IEEE Pervasive
Computing, Dec 2003)
Some energy saving strategies:
• Energy harvesting: Energy from vibrations, charging mats, moving humans
• Scale content: Reduce image, video resolutions to save energy
• Auto-dimming: Dim screen whenever user not using it. E.g. talking on phone
• Better user interface: Estimate and inform user how long each task will take
E.g: At current battery level, you can either type your paper for 45 mins, watch video for 20 mins, etc
Ubiquitous Computing
Ubiquitous Computing
• Collection of active specialized assistants to assist human in tasks (reminders, personal assistant, staying healthy, school, etc)
• App figures out user’s current state, intent, assists them
• How? array of active elements, sensors, software, Artificial intelligence
• Extends mobile computing and distributed systems (more later)
• Note: System/app initiates activities, has intelligence
• Example: Google Assistant, feed informs user of• Driving time to work, home
• News articles user will like
• Weather
• Favorite sports team scores, etc
• Also supports 2-way conversations
User Context
Imagine a genie/personal assistant who wants to give you all the “right information” at the right time
Without asking you any questions
Examples:
Detect traffic ahead, suggest alternate route
Bored user, suggest exciting video, etc
Genie/personal assistant needs to passively detect user’s:
Current situation (Context)
Intention/plan
Smart Assistant/speaker- User asks questions
- Answer questions, user requests
- Stream music, order a pizza,
- Weather, news, control smart
home
VisionCurrent
“personal assistant”
Ubicomp Senses User’s Context
Context? Human: motion, mood, identity, gesture
Environment: temperature, sound, humidity, location
Computing Resources: Hard disk space, memory, bandwidth
Ubicomp example:
Assistant senses: Temperature outside is 10F (environment sensing) + Human plans to go work (schedule)
Ubicomp assistant advises: Dress warm!
Sensed environment + Human + Computer resources = Context
Context-Aware applications adapt their behavior to context
Sensing the Human
Environmental sensing is relatively straight-forward• Use specialized sensors for temperature, humidity, pressure, etc
Human sensing is a little harder (ranked easy to hard) When: time (Easiest)
Where: location
Who: Identification
How: (Mood) happy, sad, bored (gesture recognition)
What: eating, cooking (meta task)
Why: reason for actions (extremely hard!)
Human sensing (gesture, mood, etc) easiest using cameras
Research in ubiquitous computing integrates location sensing, user identification, emotion sensing, gesture recognition, activity
sensing, user intent
5 W’s + 1 H
Sensor
Example: E.g. door senses only human motion, opens
Sensor: device that can sense physical world, programmable, multi-functional for various tasks (movement, temperature, humidity, pressure, etc)
Device that can take inputs from physical word
Also includes camera, microphone, etc
Ubicomp uses data from sensors in phone, wearables (e.g. clothes), appliances, etc.
(courtesy of MANTIS
project, U. of Colorado)
RFID tags Tiny Mote Sensor,
UC Berkeley
Ubiquitous Computing: Wearables
Ubiquitous Computing: Wearable Sensors for Health
UbiComp: Wearables, BlueTooth Devices
Body Worn
Activity Trackers
Bluetooth
Wellness
Devices
External sources of data for smartphone
Definitions: Portable, mobile & ubiquitous computing
Distributed Computing
Computer system is physically distributed
User can access system/network from various points.
E.g. Unix cluster, WWW
Huge 70’s revolution
Distributed computing example: WPI students have a CCC account
Log into CCC machines,
Web surfing from different terminals on campus (library, dorm room, zoolab, etc).
Finer points: network is fixed, Human moves
Portable (Nomadic) Computing
Basic idea:
Network is fixed
device moves and changes point of attachment
No computing while moving
Portable (nomadic) computing example: Mary owns a laptop
Plugs into her home network,
At home: surfs web while watching TV.
Every morning, brings laptop to school, plug into WPI network, boot up!
No computing while traveling to school
Mobile Computing Example
Continuous computing/network access while moving, automatic reconnection
Mobile computing example: John has SPRINT PCS phone with web access, voice, SMS messaging.
He runs apps like facebook and foursquare, continuously connected while walking around Boston
Finer points: John and mobile users move
Network deals with changing node location, disconnection/reconnection to different cell towers
Ubiquitous Computing Example
Ubiquitous computing: John is leaving home to go and meet his friends. While passing the fridge, the fridge sends a message to his shoe that milk is almost finished. When John is passing grocery store, shoe sends message to glasses which displays “BUY milk” message. John buys milk, goes home.
Core idea: ubiquitous computing assistants actively help John
SmartPhone Sensing
Smartphone Sensing
Smartphone used to sense human, environment
Example: Human activity sensing (e.g. walking, driving, climbing stairs, sitting, lying down)
Example 2: Waze crowdsourced traffic
Sensor Processing
Machine learning commonly used to process sensor data Action to be inferred is hand-labelled to generate training data
Sensor data is mined for combinations of sensor readings corresponding to action
Example: Smartphone detects user’s activity (e.g. walking, running , sitting,) by classifying accelerometer sensor data
What Can We Detect/Infer using Smartphone Sensors
Image Credit: Deepak Ganesan, UMass
Smartphone Sensing!!
Smartphone
Sensor data
Machine
Learning
Internet of Things (IoT)
IoT: Definitions
Internet extended to connect Devices
New technology paradigm
Internetworked smart machines and devices can Interacting with each other
Exchanging information
Can be controlled over the InternetLee, I. and Lee, K., 2015. The Internet of Things (IoT): Applications, investments, and challenges for enterprises. Business Horizons, 58(4), pp.431-440.
IoT: Networked Smart Things (Devices)
Smart things: Can be accessed, controlled over the network, learns users patterns
Nest Smart thermostat
- Learns owners manual settings
- Turns down heat when not around
Smart Fridge
- See groceries in fridge from anywhere
Other Ubicomp Systems Smart Homes: ambient intelligence, sensing, context-aware
services, enable remote home control Alam, M.R., Reaz, M.B.I. and Ali, M.A.M., 2012. A review of smart homes—Past, present, and future. IEEE trans sys. man, and cybernetics, 42(6), pp.1190-1203.
Example: Falls kill many old people who live alone
Smartphone continuously monitors elders living in smart home, automatically dials 911 if elder falls or ill
Smart buildings: intelligently improve comfort and energy efficiency
Wang, Z., et al (2012a), “Integration of plug-in hybrid electric vehicles into energy and comfort management for smart building”, Energy and Buildings, Vol. 47, pp. 260-266.
Senses presence of people, ambient temperature, people flow, dynamically adjusts heating/cooling
Up to 40% savings energy bill
Other Ubicomp Systems
Smart Cities: intelligently improve citizens’ quality of life, transport, traffic management, environment, economy and interaction with government
Ismagilova, E., Hughes, L., Dwivedi, Y.K. and Raman, K.R., 2019. Smart cities: Advances in research—An information systems perspective. International Journal of Information Management, 47, pp.88-100.
Example: About 30% of traffic jam caused by people hunting for parking
Real time data from Sensors embedded in street used to direct drivers to empty parking spots
References
Android App Development for Beginners videos by Bucky Roberts (thenewboston)
Ask A Dev, Android Wear: What Developers Need to Know, https://www.youtube.com/watch?v=zTS2NZpLyQg
Ask A Dev, Mobile Minute: What to (Android) Wear, https://www.youtube.com/watch?v=n5Yjzn3b_aQ
Busy Coder’s guide to Android version 4.4
CS 65/165 slides, Dartmouth College, Spring 2014
CS 371M slides, U of Texas Austin, Spring 2014